to 0.89 for various grade and socio

to 0.89 for various grade and socioeconomic groups. The investigators concluded that
their test “appears to be a usable diagnostic instrument.” The major limitation they
identified with their study was “the lack of external criteria for determining the validity
of the test items, and the validity of the identified skills as a part of the problem-solving
process.” Still their research suggests fruitful avenues for others to pursue.
Butts[ 161 developed an X-35 Test of Problem Solving (forms A and B) to assess these
behaviors:
( I ) Early formation of a hypothesis.
(2) Specific experimentation with relevant variables as contrasted to random guessing.
(3) Introduction of control to test the validity of a hypothesis selected.
(4) Specific attempts at verification of the hypothesis.
The testing situation placed the student in as natural a problem situation as possible in
which the student could “select the kinds and amounts of information he believed would
best enable him to solve the problem.” Via a “tab” format, the student pulls the tab on
all items that are considered helpful in solving a specific problem. The available data
were catagorized as:
(i)
(ii) Additional or extra information.
(iii) Duplicate information.
(iv) Irrelevant information.
Once a “tab” has been pulled, it cannot be replaced, so the examiner has a record of
which items were used and in what sequence. The student responses were evaluated by
three professors based on these judgments:
to the problem solution?
variables to the problem?

(C) Did the student introduce any controls into his thinking to test his hypothesis?
On a scale of one to five, the judges evaluated each student’s responses with respect to
those four questions producing scores ranging from 4 to 20. A score of four indicates little
or no evidence of this problem-solving methodology in their thinking, while a score of
20 indicates definite evidence of such structuring and thinking. Butts cited agreement
between the evaluation of the investigator and the judges as evidence of construct validity
of the test. Interpreting the two forms of the test as halves of one test, Butts correlated
student scores from both forms to obtain a reliability coefficient of 0.54. As an innovative
attempt in a very complex domain, this test should be scrutinized closely.
Secondary School Level
The Process of Science Test (POST) [ 171 formerly called the Impact Test, is composed
of 40 four-choice items designed by the Biological Science Curriculum Study (BSCS)
to measure the “ability of students to recognize adequate criteria for accepting or rejecting
hypotheses, and to evaluate the general structure of experimental design in science, including
the need for controls, repeatability, adequate sampling, and careful measurement.”
The POST was to be one phase of the BSCS evaluation program. Items on the
POST are framed in biological science settings, but the authors claim knowledge of
biology is not a prerequisite for scoring high on the test. Many of the items are based on
tabular or graphical presentations of data or sketches of experimental setups. It was
developed to be used both as a pre- and a posttest with biology or other classes in which
the processes of science are important objectives. The test manual includes norms, reliability
information, and correlations with measures of mental ability. The present form,
copyrighted in 1963, was administered to more than 28,000 students at the beginning
of the 1962- 1963 school year and to 24,000 at the end of the school year. Generally, the
last 20 items are of lower quality when compared with the excellent quality of the first
20 items. A major weakness of the POST is that it lacks a precise table of specifications
or categorization of items as to specific processes of skills. Despite this weakness, the
POST is one of the few standardized tests in this area for secondary level students.
More recently, Tannenbaum[ 181 developed an “instrument to assess achievement
and diagnose weaknesses in the use of scientific processes by students in grades seven,
eight, and nine” entitled the Test of Science Processes (TOSP). The test is based on these
processes: Observing, Comparing, Quantifying, Classifying, Measuring, Experimenting,
Inferring, and Predicting. These processes were chosen after consulting the relevant
literature, so the dependence on the SAPA model is understandable. The TOSP has 96
five-choice items which require 73 minutes of total testing time-normally distributed
in two separate sittings. The first 12 items are based on 35 mm color slides and the large
majority of the remaining items are accompanied by pictures or data. In the validation
study, the TOSP was administered to over 3,600 students chosen to represent all ability
levels and a wide range of SES backgrounds. The KR-20 reliability of the total test with
the entire sample was 0.91. Reliability estimates of the subtests (for each of the processes)
vary from 0.30 to 0.78 with the subtest c

Results (Indonesian) 1:

to 0.89 for various grade and socioeconomic groups. The investigators concluded thattheir test “appears to be a usable diagnostic instrument.” The major limitation theyidentified with their study was “the lack of external criteria for determining the validityof the test items, and the validity of the identified skills as a part of the problem-solvingprocess.” Still their research suggests fruitful avenues for others to pursue.Butts[ 161 developed an X-35 Test of Problem Solving (forms A and B) to assess thesebehaviors:( I ) Early formation of a hypothesis.(2) Specific experimentation with relevant variables as contrasted to random guessing.(3) Introduction of control to test the validity of a hypothesis selected.(4) Specific attempts at verification of the hypothesis.The testing situation placed the student in as natural a problem situation as possible inwhich the student could “select the kinds and amounts of information he believed wouldbest enable him to solve the problem.” Via a “tab” format, the student pulls the tab onall items that are considered helpful in solving a specific problem. The available datawere catagorized as:(i)(ii) Additional or extra information.(iii) Duplicate information.(iv) Irrelevant information.Once a “tab” has been pulled, it cannot be replaced, so the examiner has a record ofwhich items were used and in what sequence. The student responses were evaluated bythree professors based on these judgments:to the problem solution?variables to the problem?(C) Did the student introduce any controls into his thinking to test his hypothesis?On a scale of one to five, the judges evaluated each student’s responses with respect tothose four questions producing scores ranging from 4 to 20. A score of four indicates littleor no evidence of this problem-solving methodology in their thinking, while a score of20 indicates definite evidence of such structuring and thinking. Butts cited agreementbetween the evaluation of the investigator and the judges as evidence of construct validityof the test. Interpreting the two forms of the test as halves of one test, Butts correlatedstudent scores from both forms to obtain a reliability coefficient of 0.54. As an innovativeattempt in a very complex domain, this test should be scrutinized closely.Secondary School LevelThe Process of Science Test (POST) [ 171 formerly called the Impact Test, is composedof 40 four-choice items designed by the Biological Science Curriculum Study (BSCS)to measure the “ability of students to recognize adequate criteria for accepting or rejectinghypotheses, and to evaluate the general structure of experimental design in science, includingthe need for controls, repeatability, adequate sampling, and careful measurement.”The POST was to be one phase of the BSCS evaluation program. Items on thePOST are framed in biological science settings, but the authors claim knowledge ofbiology is not a prerequisite for scoring high on the test. Many of the items are based ontabular or graphical presentations of data or sketches of experimental setups. It wasdeveloped to be used both as a pre- and a posttest with biology or other classes in whichthe processes of science are important objectives. The test manual includes norms, reliabilityinformation, and correlations with measures of mental ability. The present form,copyrighted in 1963, was administered to more than 28,000 students at the beginningof the 1962- 1963 school year and to 24,000 at the end of the school year. Generally, thelast 20 items are of lower quality when compared with the excellent quality of the first20 items. A major weakness of the POST is that it lacks a precise table of specificationsor categorization of items as to specific processes of skills. Despite this weakness, thePOST is one of the few standardized tests in this area for secondary level students.More recently, Tannenbaum[ 181 developed an “instrument to assess achievementand diagnose weaknesses in the use of scientific processes by students in grades seven,eight, and nine” entitled the Test of Science Processes (TOSP). The test is based on theseprocesses: Observing, Comparing, Quantifying, Classifying, Measuring, Experimenting,Inferring, and Predicting. These processes were chosen after consulting the relevantSastra, sehingga ketergantungan pada SAPA model dimengerti. TOSP memiliki 96lima-pilihan item yang memerlukan 73 menit total pengujian waktu-biasanya didistribusikandalam dua sittings terpisah. Item pertama 12 Berdasarkan 35 mm warna slide dan besarmayoritas sisa barang disertai dengan gambar atau data. Di validasistudi, TOSP telah diberikan kepada lebih dari 3.600 siswa yang dipilih untuk mewakili semua kemampuantingkat dan rentang lebar SES latar belakang. Keandalan KR-20 tes total denganseluruh adalah 0.91. Keandalan perkiraan subtests (untuk masing-masing proses)bervariasi dari 0,30 0.78 dengan subtest c